EASTERBROOK, Circuit Judge.

Pearson-Thieman No. 1-12A (No. 12A) is an oil well in Manistee County, Michigan. It taps a pool of oil trapped by a coral reef formed during the Middle Silurian (Niagaran) Period of the Paleozoic Era, when much of Michigan was under the ocean. The northwest of Michigan's lower peninsula contains a barrier reef and many isolated pinnacle reefs; southern Michigan shows a similar pattern. See K.J. Mesolella, J.D. Robinson, L.M. McCormick A.R. Ormiston, Cyclic Deposition of Silurian Carbonates and Evaporites in the Michigan Basin, 58 Am. Ass'n of Petroleum Geologists Bull. 34 (1974); Robert H. Shaver Jack A. Sunderman, Silurian Reef and Interreef Strata as Responses to a Cyclical Succession of Environments, Southern Great Lakes Area, 1 Field Trips in Midwestern Geology 139 (1983). Substantial drilling in the northern formations began in 1972, and in 1984 alone more than 300 wells were completed. Garland D. Ellis, Michigan's Silurian Oil and Gas Pools, Report of Investigation No. 2, Michigan Dep't of Natural Resources, Geological Survey Division (1978); Michigan's Oil and Gas Fields, Rept. SS-OG-84, Michigan Dep't of Natural Resources, Geological Survey Division (1984), Table 18 from 1988 Supplement. Figure 1 shows the distribution of oil-bearing reefs in Michigan, and Figure 2 shows a cross-section of the formation.

In March 1985 the plaintiffs bought a 2% working interest in No. 12A for $310,000, together with a 2% interest in the nearby Thieman-Pearson No. 1-12. At the time No. 12A was producing about 300 barrels of oil and 400 mcf of natural gas per day. It had been producing oil at that rate since its completion in November 1983, and the output of gas had increased from a starting point near 285 mcf daily. Within a few months of the sale, however, production at No. 12A began to decline and by late 1985 hovered near 170 barrels per day. In early 1986 the buyers demanded rescission. When their demands were rebuffed they filed this suit under § 12(2) of the Securities Act of 1933, 15 U.S.C. § 77 l (2); § 10(b) of the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), and the SEC's Rule 10b-5; and state laws. The state-law counts rest on pendent jurisdiction, because both the buyers and the sellers are citizens of Illinois.

The sale of the 2% interests was not registered, but the plaintiffs do not protest on that account. The complaint does not reveal whether the sellers relied on Regulation D (in particular, a Rule 506 sale to accredited investors), directly on the § 4(2) exemption for private placements, 15 U.S.C. § 77d(2), or perhaps on the implied "§ 4(1 1/2) exception" for aftermarket sales of privately-placed securities to knowledgeable buyers. (The complaint does not even make clear whether the sellers are issuers or underwriters; we know only that the working interests came via a partnership.) No matter, the plaintiffs' acquiescence in the sale without registration means that we must assume that they met the requirements of a § 4(2) or Rule 506 sale, the most important for current purposes being that plaintiffs are sophisticated, informed, and probably wealthy investors, "those who are shown to be able to fend for themselves", SEC v. Ralston Purina Co., 346 U.S. 119, 125, 73 S.Ct. 981, 984, 97 L.Ed. 1494 (1953). The name of one plaintiff, Acme Propane, suggests that it has some knowledge of the subject. At all events, the complaint does not portray the buyers as tenderfeet.

The nub of the complaint is fraud. The plaintiffs say that the sellers told them orally that No. 12A had a production history "identical" to that of No. 12 and would have an identical productive life: that both wells "would produce oil for thirty years". These representations are false, plaintiffs say; moreover, plaintiffs insist that the defendants withheld information showing that the ratio of gas to oil produced by No. 12A was rising (which plaintiffs portray as a dire sign); that the choke orifice of No. 12A had been enlarged (ditto), and that Michigan limits the gas production from any well to 450 mcf daily.

The defendants filed a motion under Fed.R.Civ.P. 12(b)(6) to dismiss the complaint for failure to state a claim on which relief may be founded. The district judge granted the motion, 666 F. Supp. 143. The court did not (and given the posture of the case could not) dispute that the defendants made the statements attributed to them, and that the statements are false. It held only that the defendants revealed the truth about No. 12A in a three-page "Reserve Estimate" that had been provided to the plaintiffs. Relying on Teamsters Local 282 Pension Trust Fund v. Angelos, 762 F.2d 522, 529-30 (7th Cir. 1985), the court concluded that deceitful statements are not "material" to the investment decision when the investor has been told the truth; one cannot be deceived by what one knows is false. That dispatched the claims under federal law, and the district court relinquished its pendent jurisdiction of the state claims.

Zobrist v. Coal-X, Inc., 708 F.2d 1511 (10th Cir. 1983), deals with a stark contradiction between oral and written disclosures and holds that the written disclosures control. Angelos indicates this court's inclination to follow Zobrist, on two grounds. First, only "material" misstatements permit recovery under the securities laws, and to be material a statement must significantly alter the total mix of information available to the investor. Basic Inc. v. Levinson, ___ U.S. ___, 108 S.Ct. 978, 983, 99 L.Ed.2d 194 (1988). An investor who knew the truth is hard put to say that the inconsistent oral statement significantly altered the total mix of information. Second, the securities laws are designed to encourage the complete and careful written presentation of material information. A seller who fully discloses all material information in writing should be secure in the knowledge that it has done what the law requires. Just as in the law of contracts a written declaration informing one party of an important fact dominates a contrary oral declaration, so in the law of securities a written disclosure trumps an inconsistent oral statement. Otherwise even the most careful seller is at risk, for it is easy to claim: "Despite what the written documents say, one of your agents told me something else." If such a claim of oral inconsistency were enough, sellers' risk would be greatly enlarged. All buyers would have to pay a risk premium to cover this extra cost of doing business.

Zobrist and Angelos dispose of several of the buyers' contentions. For example, the buyers say that the sellers did not disclose the increasing ratio of gas to oil produced from No. 12A, which, the buyers say, presages appearance of a "gas cap" at the producing horizon, with baleful consequences. The material available to us suggests that an increasing ratio of gas to oil is characteristic of both wells with free gas caps and wells with solution gas drives, and that an increase early in production is more likely to indicate that the well is being driven by gas in solution (which the plaintiffs say is what they were hoping for at No. 12A). Figure 3 is from A.I. Levorsen, Geology of Petroleum 443-45 (1956).

We need not decide whether this is the sort of information of which we may take judicial notice. No matter what we make of the rising ratio at No. 12A, however, the sellers revealed the information to the buyers in writing. They furnished tables of the monthly production of oil and gas from Nos. 12 and 12A since each had been completed. Long division would have produced the ratio of one to the other.

These tables also squelch any claim that the sellers misled the buyers into thinking that the two wells are identical. The tables reveal them to be dissimilar, with No. 12 producing about 100 barrels of oil and 400 mcf of gas per day at the time of the sale, and No. 12A about 300 barrels and 400 mcf. The gas-oil ratio at No. 12 was higher and had fluctuated substantially. The difference between the wells seemed to run in favor of No. 12A. As for the claim that the sellers stated orally that there had been no change in the diameter of the choke orifice: page 1 of the Reserve Estimate states that "no appreciable decline in pressure has been observed that cannot be attributable to changes in the choke setting orifice." A three-page Reserve Estimate is not so long that such a statement is too obscure to find.

The Reserve Estimate did not reveal that Michigan sets a limit of 450 mcf of gas daily, but no oral misrepresentation needed to be set right. Production limits are designed to increase the portion of any hydrocarbon pool that can be extracted; every state with oil and gas deposits uses them; the failure to disclose the particular nature of Michigan's cannot have caused the buyers any loss. The limit is not specific to No. 12A. The securities laws require the disclosure of information that is otherwise not in the public domain. Cf. LHLC Corp. v. Cluett, Peabody Co., 842 F.2d 928, 931 (1988). Sellers of securities need not "disclose" the statutes at large of the states in which they operate, any more than they have to disclose how the acquisition of oil for the Strategic Petroleum Reserve may affect the value of a 2% share of an oil well.

Then there is the oral statement that No. 12A would produce for thirty years. This could have at least three meanings: (1) "No. 12A will produce something for 30 years" — a statement trivially true, because any well can be made to produce until it rusts away by curtailing the rate of production. (2) "No. 12A will produce at its current level for 30 years" — a statement trivially false, because only a freak well maintains its initial level for that long, and no modestly sophisticated buyer (which we must suppose the plaintiffs to be) could give the 30-year statement this meaning. (3) "No. 12A will produce at a commercially satisfactory rate for 30 years." This statement could be both material and false, and it is the sort of falsity that could escape immediate detection by a modestly sophisticated buyer. Some wells do last that long, including some in Michigan. Michigan's Oil Gas Fields at 243; Ellis, Silurian Oil Gas at 29-40. Because this case arises on motion to dismiss, we must use the third reading, the plausible one most favorable to the plaintiffs.

This reading interacts with the statement that the production histories of Nos. 12 and 12A are identical. We have held not actionable the reading of this statement as "They produce identical amounts daily"; information in the plaintiffs' hands refuted that interpretation. But a reasonably sophisticated buyer might read this statement as meaning that Nos. 12 and 12A tap the same pool of oil, so that they will behave consistently through time; or as saying that their reservoirs are the same size; or as saying that they will behave with equal consistency through time. (That is, that they are both stable producers; No. 12 had been producing 80-100 barrels daily since January 1982, with small fluctuations.)

Any of these statements would be material, and any could be false — as (ex post) they have proved to be. (They are actionable, of course, only if their falsity was known or knowable in March 1983, something we are assuming for the sake of argument.) No. 12A's production dropped sharply in 1985 (we do not have subsequent data), while No. 12's production stayed constant. Moreover, a measure of reservoir size in May 1985 implies that No. 12A's reservoir is about half the size of No. 12's. A good way to estimate the size of a reservoir is to measure the gas pressure at the bottom on two occasions (these are called "shut-in tests"). The operator knows how much oil was extracted between the two dates and learns how far the gas pressure has fallen. This permits him to compute how much oil was extracted for each lost pound of pressure; and since he knows how much gas pressure there was to start with he can compute roughly how much oil is commercially extractable in the primary phase of production. Data on No. 12 show that it is producing about 300 bbl/psi drop in pressure and started with 1990 psi at the bottom of the hole. Subsurface Pressure Report on the Northern Michigan Reef Development, Michigan Dep't of Natural Resources, Geological Survey Division, Sec. I (1984 1986). When No. 12A was completed, it had a bottom-hole pressure of 2293 psi. A test on May 6, 1985, revealed a bottom-hole pressure of 1325 psi, implying about 150 bbl/psi drop. Subsurface Pressure Report Sec. II. Since No. 12A started with only 300 psi higher pressure than No. 12, this implies that the primary recovery from the pool under No. 12 will be almost twice that from the pool under No. 12A.

The written Reserve Estimate disclosed that the operator had not performed a second shut-in test, and so did not yet have this estimate of the reservoir at No. 12A. Nothing in the written Reserve Estimate is inconsistent, however, with several plausible implications of the oral statements: that No. 12A would produce at a commercially reasonable rate for 30 years, that it would be as stable as No. 12, that it would produce as much as No. 12 (maybe even three times as much, the ratio in March 1983). Quite the contrary, the Reserve Estimate was bullish on No. 12A. It estimated that No. 12 would yield about one million barrels and continued:

Armed with this knowledge, and the history of similar profilic [sic] wells in the area we can only assume that this well [No. 12A] will perform equally as good if not better. To quote Don Yohe, the operator of the lease, "of course my interest is not for sale but if it was I wouldn't take less than sixty (60) times the monthly income. Heck it's worth every bit of that."

In other words, the Reserve Estimate used No. 12 and other nearby wells (which it characterized as similar) as the basis of a projection that the reservoir under No. 12A contained more than one million barrels of oil.

That is not a truth that cured any falsehood. It is closer to a repetition of the statements alleged to be false. Worse, it may contain deception of its own. The sellers invited the buyers to infer the productivity of No. 12A from the most productive surrounding wells. But from everything we can find in the published literature about reefs in Manistee County, the history of one well is worthless as an indicator of the behavior of nearby wells completed in other pools. Each reef is unique; size and other pertinent properties such as permeability are unpredictable and very hard to measure until at least two shut-in tests have been performed. Yet a statement such as "The past production at Well A is a good indicator of the future production at Well B" could mislead even a sophisticated buyer, because the statement would be true if the production came from a larger isotropic reservoir (underlying both wells) or perhaps even from barrier reefs rather than from pinnacle reefs. But the only evidence available shows that No. 12 and No. 12A were completed in noncommunicating reefs. Shut-in tests performed within a month's interval at No. 12, No. 12A, and in the other pool mentioned in the Reserve Estimate show a variation of more than 1200 psi at producing horizons near the same depth. Subsurface Pressure Report Sec. II. This means that the wells were completed in different reefs, or at least in different pools in the reef complex, with no reason at all to expect similar production characteristics. As we must take this record, the use of No. 12 as the basis for predicting the future production of No. 12A had no better basis than the use of divining rods to measure the size of pools. Cf. Rule 175(a). The Reserve Estimate, far from correcting any errors in the oral statements, may contain an actionable misrepresentation of its own.

Angelos and Zobrist hold that written words govern oral ones in order to reward truthful disclosures and facilitate accurate assessment of risk. The rule of those cases is not designed to make it easy to perpetrate oral frauds. We keep in mind the need to deter deceit, both oral and written, and so we shall insist that the written words be true, clear, and complete, in order to be dispositive. The Reserve Estimate is none of these.

All we have said is based on the record in its current state, which includes only what the buyers attached to their complaint and published documents about oil-producing formations in Michigan. The sellers may be able to cast a completely different light on this. But in disposing of a motion under Rule 12(b)(6), the district court must give the plaintiff credit for everything that might reasonably be proved in support of the complaint. Adams v. Bell, 711 F.2d 161, 187 (D.C. Cir. 1983). This complaint presents some allegations that, if proved to have been made, and made falsely, are material and not contradicted by truths in the Reserve Estimate. The judgment is therefore reversed, and the case is remanded for further proceedings consistent with this opinion.